Chemical interactions in the aggregation of bacteria bioflocculation in

ting that bacteria can be flocculated with synthetic anionic and nonionic polyelectrolytes (polyacrylamide, polystyrene sulfonate, polyglutamic acid, ...
0 downloads 0 Views 3MB Size
Chemical Interactions in the Aggregation of Bacteria Bioflocculation in Waste Treatment Paul L. Busch and Werner Stumm

Division of Engineering and Applied Physics, Harvard University, Cambridge, Mass. 021 38

Bacteria Aggregation of microorganisms is effected by an interaction of polymers excreted by the microbial cell or exposed at their surface. This proposition has been tested by demonstrating that bacteria can be flocculated with synthetic anionic and nonionic polyelectrolytes (polyacrylamide, polystyrene sulfonate, polyglutamic acid, and dextran). As with hydrophobic sol-polymer systems with charges of equal sign, reduction of surface potential is not a prerequisite for flocculation; agglomeration apparently results from specific adsorption of polymer segments and from bridging of polymers between the cells. Extracellular or cell surface polymeric material isolated from cell-free supernatant of a well flocculating activated sludge culture is, similar to synthetic polyelectrolytes, capable of destabilizing dispersions of bacteria and of abiotic sols.

any microorganisms have a tendency to adhere to interfaces and to each other; tinder certain conditions trey remain freely dispersed ; under others they form aggregates. Microbial aggregation influences the temporal and spatial distribution of nutrients in aquatic habitats. Flocculation is of significance in biological waste treatment. The efficiency of the conventional activated sludge process greatly depends upon bioflocculation and subsequent phase separation of the flocs from the treated elfluent. Natural polymers are of great importance in sorptive and colloidal destabilization reactions in biological systems. For example, the reaction between antibody and antigen (Goldberg, 1952; Katchalsky, Danon, et a/., 1959), the cohesion of tissue cells (Pethica, 1961; Moscona, 1962), the aggregation of sponge cells (Humphreys, 1965), the flocculation of yeast cells (Mill, 1964), and the stabilization of soil aggregates (Greenland, Lindstrom, et a/., 1961) have been interpreted in terms of colloid-polymer interactions. Similarly, Tenney and Stumm (1965) have proposed that flocculation of microorganisms in biological waste treatment is effected by an interaction of polymers excreted by the microorganism or exposed at the microbial surface under suitable physiological conditions. According to this hypothesis, flocculdtion of the dispersed organisms is interpreted as a process whereby polymers adsorb to and bridge between cell surfaces. This research was undertaken to provide experi,zxniai corroboration for a polymer mechanism in bioflocculation and a better understanding of some of the factors that influence aggregation of microorganisms and other colloids in biological waste treatment.

cis

Hydropliilic Colloids

Within the pH-range of interest (pH 5 to 9), bacteria carry a net negative charge resulting from acid-base ionization of functional groups in the bacterial surface. Because of the hydrophilic surface, the stability of a bacterial dispersion does not primarily depend on the electrostatic repulsive forces between the cells. Reduction of surface potential is not a prerequisite for bioflocculation; bacterial suspensions can form stable dispersions even at the isoelectric point (pH 2 to 3 for most bacteria). Electrolyte concentrations capable of coagulating hydrophobic sols have no efect in destabilizing bacteria. The cell-free supernatant of a bacterial culture. presumably containing material excreted bq the organisms or sheared off the cell surface, is capable of imparting a negative electrophoretic mobility to incipiently positively charged A120j. It may be inferred that polymeric material excreted or exposed at the surface of microorganisms contains functional groups that are primarily anionic and nonionic in the neutral pH range. Therefore, if aggregation of bacterial cells is assumed to result from polymer interactions, the anionic and nonionic polymeric cell products must be capable of interacting with the negatively charged bacterial surface. To test the proposition that specific chemical interactions can occur between bacterial surfaces and polymers which are not opposite in charge characteristics. i t appeared desirable to validate experimentally that microorganisms can be flocculated by synthetic anionic and nonionic polyelectrolytes and to show that extracellular polymeric material produced by the cell has flocculative properties similar to those of synthetic polyelectrolytic flocculants. Mtrter.i(ils iiticl Metliods

Polyacrylamide, polystyrene sulfonate, and polygalacttironic acid and dextran were supplied by American Cyanamid. Dow Chemical Co.. and Sigma Chemical Co.. respectively. The bacteria used for pure culture experiments were obtained from the American Type Culture Collection. The methods employed in this investigation have been described in detail by Busch (1966). A continuous supply of bacteria of similar physiological state and similar surface characteristics was maintined by a 500-ml. single-stage continuous culture apparatus, using incipientlk sterile media containing glucose or benzoate iis carbon sources. Dilution rates in most experiments were between 0.5 and 0.8 hourVarious methods of measuring bacterial aggregation (counting microscopically or electronically, settling, cake filtration, and light scattering) have been compared (Btiscli, 1966); Volume 2, Number 1, Januar? 1968

49

no method tested was entirely adequate; light scattering intensity, r , (the ratio of the intensity of light scattered at 90" to the light transmitted) was considered most satisfactory although not always sensitive. While not quantitative, light scattering results do present a qualitative measure of aggre gation while requiring no sample dilution or extensive measurement times. A reduction in r indicates an increase in aggregation at constant weight concentration of particles. For comparative purposes the relative scattering intensity 7 / 1 0 is used to indicate a change in aggregation; 70 is the light scattering intensity of a reference sample--e.g., sample without polymer added. Measurements of r were made at 546 mp (Brice Phoenix light-scattering photometer) 20 minutes after addition of polymer and subsequent slow stirring of the bacterial-polymer dispersion. Material containing cell-free flocculants was isolated from the supernatant of a well flocculating activated sludge culture maintained in the laboratory. The bacteria in this culture were concentrated by gravity sedimentation for 1 to 2 hours. The heavy sludge that resulted was centrifuged a t ca. 24,000 G. The supernatant from this centrifugation was added to an ethanol solution (3 parts ethanol for 1 part supernatant). The precipitate which developed was centrifuged and resuspended in an ethanol-water solution (3 to 1) three times and finally air-dried.

Results and Discussion

Flocculation of Bacteria with Synthetic Anionic and Nom ionic Polyelectrolytes. Several representative flocculation curves describing the aggregation of Aerobacter aerogenes or Escherichia coli with various polyelectrolytes are given in Figure 1. These curves are typical; at polymer concentrations larger than those required for optimum flocculation, redispersion or restabilization of the bacterial suspension occurs. As illustrated in Figure la, the electrophoretic mobility, reflecting a negative surface potential on the bacterial interface, becomes more negative as a result of the addition and subsequent adsorption of the anionic polyelectrolyte. Similar experiments were also conducted with polyglutamic acid and with dextran. Polyglutamic acid produced flocculation a t a pH of ca. 4. At this pH, the polyglutamic acid molecule has a net charge of zero. Bacterial aggregation with nonionic dextran (M.W. = 6,000,000)occurs, however, with very little efficiency a t a pH optimum of approximately 5. The forces responsible for attachment of the polyelectrolyte segments to the bacterial surfaces appear to be relatively weak, since increased agitation (stirring a t >30 r.p.m. with magnetic stirring bars) results in shear forces that disperse the aggregates. The pH of the medium has a rather pronounced effect on the efficiency of the aggregation. Different polymers require different pH optima; no satisfactory explanation for these p H effects can be given. The phenomena involved in the flocculation of colloidally dispersed hydrophobic solids by polyelectrolytes has been interpreted satisfactorily as a process of specific adsorption of polymers or polymer segments on the colloidal surface 50 Environmental Science and Technology

I

02c

P 3 rlClylormde

MU=XU,W,

wOtiyar

lmql~1

"

~ ~ y q o ~ o c t m i nc lca

e

1mq/-l

Figwe 1. Aggregation and redispersion of bacteria by anionic and lloldonic polymers qb,e. Anionic polyelectro4tes c. Nonionic polysaccharide d. Polyglutamic acid has no net charge at pH-4

leading to bridging of polymer chains between solid particles (Ruehrwein and Ward, 1952; Michaels, 1954; Smellie and LaMer, 1958). The effect of polyelectrolytes on hydrophilic colloids such as bacteria may be explained similarly. In accordance with such a model, optimum destabilization occurs when a certain fraction of available adsorption sites on the surface of the bacteria are bridged by polymers. Correspondingly, the optimum polymer dose should be proportional to the colloid concentration, or more exactly, to the available suiface area. Such a linear relationship has been observed in this investigation (Figure 2). When polymer concentrations exceed this optimum, bridging is limited and cells become redispersed since many of the reactive surface sites are occupied by polymers, which are then unable to find unfilled sites on other colloids or are hindered by steric (Heller and Pugh, 1960) or electrostatic (LaMer and Healy, 1963) interferences between polymer segments. In this theory, polyelectrolyte segments must be held on the bacterial surface by forces that can outweigh those of purely electrostatic origin. Possible mechanisms for the attachment of anionic functional groups to negatively charged surfaces are : hydrogen bonding; anion interchange with adsorbed anions (such as OH-); or interactions with cations on or in the immediate vicinity of the bacterial surface. Flocculation of a polymer-sol system with charges of equal sign occurs only if an appropriate concentration of an electrolyte is present in the solution (Overbeck, 1938; Deveaux, 1939; Black, Birkner, et al., 1966; Sommerauer, Sussman, et al., 1967; Nemeth and MatijeviE, 1967). However, flocculation with polyelectrolytes can occur a t electrolyte concentrations much smaller than those necessary for flocculation in the absence of polymers. In these investigations, satisfactory flocculation with polymers was observed only when the concentration of divalent cations

exceeded a certain minimum (for example, 3 X lO-‘M of Caf2 or Mgf2). In the absence of polymers, bivalent ion concentrations larger than 10-lM produced aggregation. In accordance with earlier investigations on the effect of electrolytes on the stability of a hydrophobic negative sol-anionic polymer system (Soimerauer, Sussrnan, et uf., 1967), the authors believe that the primary role of divalent cations in the flocculation of bacteria is complex or ion pair formation between the functional groups of the polyelectrolyte and the “counter ion” in the immediate vicinity of the bacterial surface, thus enhancing the adsorption of polymer segments at the interface. Although concentrations of bivalent cation and polymer are not sufficiently high in the bulk solution to lead to complex formation, coordination becomes possible in the vicinity of the negatively charged bacterial surface where, for electrostatic reasons, incipient cation concentration is much larger than in the bulk of the solution. Flocculatioa with Polymers of Bacterial Origin. Material containing cell-free flocculants was isolated by ethanol precipitation from the supernatant of a centrifuged activated sludge culture fed on glucose. This precipitate presumably contained some of the capsular substances loosely bound t o the microbial surface and other natural polymers that might have been excreted by the organisms. Analysis of the precipitate revealed no appreciable quantities of nucleic acids or protein, but the material gave a positive anthrone test indicative of carbohydrates. An aqueous solution of this material was capable of imparting a negative electrophoretic mobility to AIDs particles of incipient positive surface potential a t a pH of 7 indicating that the interfacially active components in the cell free supernatant were anionic in the neutral pH range. These components are also capable of flocculating inorganic colloidal dispersions with a negative surface potential. Figure 3 shows typical flocculation-redispersion curves for silica and for A. aerogenes. An electron micrograph of a representative silica floc taken from this experiment is shown in Figure 4. The spacing between the individual particles is relatively uniform and rather large (100 to lo00 A.). The fact that similar intraparticular distances have been reported for the aggregation of cells by biological polymers (Ambrose, 1964; Humphreys, 1965; Pethica, 1961) is suggestive. The material containing the natural polymers is also capable of aggregating suspensions of dispersely growing A. aerogenes (Figure 3,b). This flocculation curve is similar to the curves describing the interactions between synthetic polyelectrolytes and negatively charged colloids. The observation that extracellular products of the organisms can change the stability of dispersed cells and inorganic sols in a similar manner suggests that these polymers may also form bridges between cells and other solids thus coflocculating inorganic colloids and bacterial cells. Figure 5 shows that the flocculative removal of dispersed silica from solution is enhanced by the addition of well flocculating bacteria.

A. aerogenes -~

II

0 30 mg/L 0 45ma/L 0 75 & / L

I

I

50

IW

150

teria (mg/L)

Figure 2. Relation between concenlration of bacteria and the optimum dosage of polyacrylamide

-12

I

Volume 2, Number 1 , January 1968 5 1

Figure 5. Floeeulating bacteria enhance the Fernoval of S i 0 2 partides

BioEmuIation. Uncertainty exists regarding the mechanism of natural bioflocculation in biological waste treatment. The various mechanisms that have been proposed have been

Figure 4. Eleelron micrograph of p d c l e s aggregated with p l y m

of bacterial origin

reviewed recently by Crabtree, Boyle, et a/.(1966). A typical biological Roc consists of a great variety of hacteria and occasionally other organisms such as protozoa, fungi, and viruses, as well as abiotic suspended matter. Thus, a Roc may comprise a broad spectrum of hydrophobic and hydrophilic interfaces. Proteins, lipides, and polysaccharides combined in various proportions occur on the surfaces of microorganisms in the form of cell walls, capsules, Ragellae, fibrils, filaments, and fimbriae ofvarious kinds. The morphological appearance of the floc depends in a complex way on the structure of the microbial population and the nutritional conditions in the medium. Bioflocculation results from complicated physical, chemical, and biological interactions; many variables, difficult to control experimentally, are involved. Different mechanisms of Roc formation may he operative for biological flocs originating from different organisms or combinations of organisms; a few strains of microorganisms may remain dispersed in the presence of excess polymer and may flocculate only after the excess has been removed. No single and unifying interpretation can be given. BioRocculation is strongly influenced by the growth phase of the microorganisms; generally, the tendency for hioflocculation is larger under conditions of declining or endogenous growth than under conditions of prolific growth. Some strains of microorganisms are known to synthesize extracellular polysaccharides or other polymers, especially under conditions where growth is limited by a nutrient other than carbon--e.g., nitrogen. The chemical transformations and condensations necessary for synthesis of polymers demand little energy and can be carried out by the cells even in the absence ofgrowth. However, since most municipal wastes are nutritionally deficient in carbon (atoms CIN < 9, other nutrients pI, P) are usually not growth limiting in biological waste treatment. Extensive bioproduction of polymers in large quantities is less likely to occur under such conditions. Nevertheless, cell binding polymeric material conducive to flocculence may also be produced under declining or endogenous carbon limited growth. Perhaps an adequate concentration of polymer can accumulate per unit surface area only under conditions of declining or endogenous growth. While organisms are growing prolifically, new microbial

surfaces are produced faster than surfaces can become covered with polymers. According t o this simple picture, newly formed cells remain dispersed as long as they replicate rapidly; thus, in the presence of sufficient metabolites, dispersed bacteria outgrow flocs. If growth becomes very slow or ceases-Le., under conditions of long detention time in continuous culture or biological treatment systems-the “old” not readily replicating cells adhere to each other and form aggregates. Furthermore, because settleable organisms are recirculated, flocculating cells are selectively retained in the activated sludge tank, while dispersed cells are washed out of the system. Practical Aspects. In aerobic biological treatment of waste, there is an incompatibility between the requirements for large specific substrate utilization rates (high growth rate, small detention times) and the requirements for maximum bioflocculation (small growth rates, large detention times). In treatment processes with submerged flocs (activated sludge), the rapid removal of organic matter during the initial contact period (the so-called contact stabilization) may, a t least t o a large part, be attributed to the flocculation of colloidal and macromolecular organic substances. Addition of suitable chemical flocculants, or control of the chemical parameters which afect natural aggregation, are among the possibilities for improving the efficiency of bioflocculation. Natural polymers isolated from well flocculating bacteria or from “slime” producing algae or other plant biocolloids may be more expedient flocculants than the synthetic polymers used in this and an earlier study (cationic flocculants; Tenney and Stumm, 1965). Conclusions

In this investigation, the authors have attempted to a p proach only the issue concerning the nature of the interactions of the cells with each other and with other interfaces. The results presented appear to confirm the general working concept that polymeric extracellular or cell-surface material is involved in bacterial aggregation. The interaction forces operative in biological self-flocculation appear to be stronger than those encountered in the aggregation by synthetic polyelectro1)teb. Bioflocs are not as sensitive to dispersion by shear forces as bacterial aggregates flocculated by synthetic polymers. Natural flocs also are not so critically dependent on pH and polymer concentration. Little is known of the chemical nature of flocculent material produced by the cell. Quite likel) different organisms produce different cell binding materials: these may be of low molecular weight and soluble or of high molecirlar weight with a high degree of crosslinkage and exist a b a rather “gelatinous slime.” The molecular frame-

work may in many cases be comprised of polymeric carbohydrates containing anionic and nonionic functional groups, such as 4 0 0 - and -OH groups. The fact that synthetic polymers containing similar anionic ( 4 0 0 - and --SO3-) and nonionic ( 4 H )groups ale able to aggregate organisms is considered a valid demonstration that natural anionic polymers may, under suitable conditions, become specifically adsorbed a t microbial surfaces and thus able to form bridges with adjacent surfaces leading to aggregates. This hypothesis is further reinforced by the observation that interfacially active anionic material precipitated from a cell-free supernatant of a flocculating activated sludge culture can destabilize dispersions of bacteria and of abiotic sols. Acknowledgment

Valuable advice given by Mark W. Tenney and Robert H. Harris and technical assistance by Margaret Flint are acknowledged. Literature Cited

Ambrose, E. J., in “Recent Progress in Surface Science,” J. F. Daneilli, K. G. A. Pankhurst, and A. C. Riddiford. Eds.. Academic Press, New York, 1964. Black, A. P., Birkner, F.. Morgan, J. J., J. Colloid lnterfirce Sci. 21,626 (1966). Busch, P. L., Ph.D. thesis. Harvard University, Cambridge. Mass., 1966. Crabtree, K.. Boyle, W.? McKay, E., Rohlich. G. A., J. Wuter Pollution Control Federation 38, 1968 (1 966). Deveaux, J., J. Chitn. Ph.vs. 36, 5 (1939). Goldberg, R. J., J . Ani. Chem. SOC.74,5715 (1952). Greenland, D. J., Lindstrom, G. R., Quirk, J. P., Narure 191, 1883 (1961). Heller, W., Pugh, T. L.,J. folwiier Sci. 47,203 (1960). Hunphreys, T., Erptl. Cell Res. 40,539 (1965). Katchalsky. A.. Danon, D.. Nevo, A.. Bioclrini. Biopliys. Actu 33, 120 (1959). LaMer, V. K., Healy, T. W.? Rer. Pure Appl. Cliem. 13, 112 (1 963). Michaels, A. S., Intl. Eng. Clieni. 46,485 (1954). Mill. P. J.. J. Cen. Microhiol. 35, 58. 61 (1964). Moscnna. A. A.. J. Cellulur Cotnp. Pli.vsiol. 60,65 (1962). Nemeth. R.. Matijevic, E.. Kolloid Z. Z. Polwiere (in press). Overbeck, J . Th. G.. Clieni. Weekhlud 35, 1178 (1938). Pethica. B. A.. Euprl. Cell Res. Suppl. 8, 123 (1961). Ruehrwein. R. H.. Ward, D. W.. Soil Sci. 73,485 (I 952). Smellie, R. H., LaMer, V. K.. J. Coilloid Sei. 13, 589 (1958). Sommerauer, H., Sussman. D., Stumm. W.. Kolloid Z. Z. Polniiere (in press). Tenney, M. W., Stumm, W.. J. Writer Polliction Conrrol Fedenition 37, 1 370 (1 965). Receice~ljiirreL-iew Airgirsr S: 1967. .4cceptet/ Decrniher 6 . 1967. Work s r ~ p ~ n w i rin i l purt 1., reseurcii grtmr W’P ~)OOYSof the Wutrr Poflrriion Control . ~ ~ ~ t l i ~ n ~ . ~ t ~ ~ i t ~ f ~ f l .

Volunie 1. Sumber I . Jantrar? 1968

53